Challenges in Achieving Height Modernization in Alaska Crustal Deformation Has Invalidated Much of the Historical Data

1
Challenges in Achieving Height Modernization in
AlaskaCrustal Deformation Has Invalidated Much
of the Historical Data

• Jeff Freymueller
• Geophysical Institute, University of Alaska
  Fairbanks

2
Height Accuracy Reaches New Low

• Terrestrial height network (leveling) is
  significantly compromised
• Last systematic leveling in southern Alaska in
  1964-1965, immediately after 1964 Prince William
  Sound earthquake
• There has been up to 1.25 METERS of vertical
  crustal motion since then.
• Substantial areas of gt20-30 cm deformation since
  last survey
• Geoid height errors in Alaska can be very large,
  as is well-known at NGS
• GRACE geoid compared to EGM-96 showed up to 1
  meter geoid height errors
• CONCLUSION Alaska needs an essentially new
  vertical reference network just like starting
  over

3
Main Sources of Vertical Motion

• Glacial-Isostatic Adjustment (post-glacial
  rebound) up to 3.5 cm per year
• 1 meter change every 30-50 years
• But signal is strongest in parts of Alaska that
  had minimal height information aside from tide
  level
• Postseismic Deformation Following the 1964
  earthquake up to 1.25 meters since 1964
• Other faulting-related deformation much smaller
  (several mm/yr)
• Deformation assoicated with 2002 Denali fault
  earthquake substantial but localized
• Rarely more than 10 cm vertical change away from
  fault
• Postseismic changes continue, may reach 10 cm
  level

4
Post-1964 Postseismic Uplift
1964 to present Units cm
5
Post-1964 Postseismic Uplift

• Estimate based on GPS surveys of leveling BMs
• Correction for geoid-ellipsoid separation is by
  far the largest source of error
• (Relative) geoid heights used, error may approach
  20 cm
• Strong gradient near Anchorage (100 cm
  mid-Turnagain Arm, 30 cm at Port of Anchorage
  tide gauge) could include a component of geoid
  error.

Data Cohen, S. C., and J. T. Freymueller,
Crustal Deformation in Southcentral Alaska The
1964 Prince William Sound Earthquake Subduction
Zone, Advances in Geophysics, 47, 1-63, 2004.
6
Tide Gauges
Linear sea level trends along strike-slip boundary
Non-linear sea level trends in transition and
subduction zones
Tectonic influence on long term uplift records is
strongest following the 1964 earthquake
Larsen, C. F., R. Motyka, J. Freymueller, and K.
Echelmeyer, Tide gauge records of uplift along
the northern Pacific-North American plate
boundary, 1937 to 2001, J. Geophys. Res.,
108(B4), doi10.1029/2001JB001685, 2003.
7
Uplift Rates 2 mm/yr contour interval
Yakutat Icefields Peak Uplift Rate 3.5 cm/year
Glacier Bay Peak Uplift Rate 2.5 cm/year
Larsen, C. F., R. J. Motyka, J. T. Freymueller,
K. A. Echelmeyer and E. R. Ivins, Rapid uplift of
southern Alaska caused by recent ice loss,
Geophys. J. Intl., 158, 1118-1133, 2004.
8
What Needs to be Done

• Existing vertical datum is in error by gt30 cm in
  much of southern Alaska, as much as 1 m in
  places.
• Because of crustal movement since surveys that
  defined it.
• May be difficult to even maintain a consistent
  definition with past
• Repeating all the leveling effectively impossible
• All southern coastal tide gauges have moved gt30
  cm since surveys used to define NAVD88
• Need new definition of vertical datum based on
  GPS
• CORS plus some number of monuments in the ground
• Need new geoid model
• If Arctic Ocean can have a 5-10 cm geoid, why
  should (populated) Alaska have to settle for
  meter-level errors?
• Its almost like starting from scratch.
• Might be better to start over with a new system
  based on ITRF. The Lower 48 will have to go that
  route some day.

9
What is Postseismic Deformation?

• Transient deformation triggered by an earthquake
• Afterslip on the fault zone
• Viscoelastic relaxation of the mantle or lower
  crust
• Poroelastic deformation associated with
  earthquake-driven fluid flow (changes elastic
  constants)

10
Characteristic Relaxation

• Postseismic deformation rate decays with time
• Viscoelastic relaxation (1-exp(-t/t)) (viscosity)
• Afterslip log(1 t/t) (frictional parameters)
• Poroelastic relaxation maybe erf(t/t) (hydraulic
  diffusivity)
• In general, all of these will be active and
  multiple timescales should be expected

11
Postseismic Uplift Rate Was Very Fast Right After
Quake
Uplift rate declined with time but still peaks at
1 cm/year